Research On Intelligent Multi-scale Representation Of Spatial Data Based On Interest Field Model

In the era of big data,various information related to geographical spaces closely are increasing largely,and the people's requirements for the cognition of spatial information is growing continuously.As customized,personalized information service has become the new trend of geographic information application,users start to play a more important role.How to use multi-source massive data to provide the flexible,efficient and personality network geographic information service has become an important subject of current research,combined with the characteristics of user's interest.This paper proposes an idea of intelligent multi-scale representation of spatial data based on interest field model.It takes the multi-source spatial data as the research subject.By analyzing the spatial distribution of the user's interest in geographic information,and combining with the characteristics of multi-scale spatial data and remote sensing image progressive coding,it establishes the user interest driven multi-source and multi-scale progressive representation technology system.It can not only realize the user's interests driven important-to-secondary knowledge level of spatial data multi-scale expression,but also increase the spatial information application level under the network environment.Thus it has important practical significance and application value.The main content of this paper are as follows:1.Based on the realistic needs of the expression and application of massive spatial data,this paper expounds the background and significance of the research,and introduces the present situation of user interest model based on the analysis of the development trend of personalized information service.Through reviewing the research progress of scale problems of geographic information,this paper analyzes the present status of multi-scale expression of spatial data,and the existing problems in its practical application.Then this paper proposes the idea of multi-scale intelligent representation of spatial data constrained by interest field,and determines the research contents and the overall research framework.2.This paper systematically introduces the basic concept and theory of multi-scale expression,analyzes the principle and features of several typical multi-scale expression technique,and then summarizes several database construction methods of the geographic elements multi-scale expression.3.This paper puts forward the concept and construction method of spatial interest field.Combing with the characteristics of big location data,it defines the concept of spatial interest field,and proposes a method of constructing spatial interest field based on POI data from the perspective of user spatial cognition and interest.Then it applies the spatial interest field model to analyze the spatial distribution features of user's interest and put forward the concept and calculation method of interest carrying capacity.After that,this paper makes a multi-scale intelligent representation scheme for spatial data constrained by interest field.4.This paper proposes the multi-scale representation of vector data and DEM data under the constraint of interest field.Using the global subdivision model to realize the spatial hierarchical decomposition and scale division of the geographical space,and combining with spatial interest field model,this paper,with the constraints of interest carrying capacity,patch interestingness and degree factor importance,presents a multi-scale representation strategy for vector maps and DEM terrain data considering user's interest,and achieves multi-scale progressive representation of multi-source geospatial data based on user interest.5.This paper proposes the multi-scale asymptotic coding of remote sensing images with interest field constraint.Dividing the region of interest with interest field,it combines Maxshift algorithm and SPIHT algorithm to code the interested region with high quality and preferentially,which realizes a progressive coding way of remote sensing image based on spatial interest field and a progressive transmission effect of remote sensing data.6.The above research results have been applied to practice,and the proposed theoretical methods have been verified by experiments.After that,the prototype system of multi-scale geospatial data intelligent expression constrained by interest field is developed.